german apprenticeship training and the skill weights …00000000-5b2d-ab7b-0000... · german...

German Apprenticeship Training and theSkill Weights Approach – An Empirical Analysis

Uschi Backes-Gellner, Johannes Mure, Regula Geel

University of ZurichInstitute for Strategy and Business EconomicsPlattenstrasse 14, CH-8032 Zurich,Tel.: + 41 44 634 42 81, Fax: +41 44 634 43 70 [email protected]

Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 2

I. The focus of our paper

- Mobility after Apprenticeship Training

- Firm`s Investment in Apprenticeship Training

II. The Skill-Weights Approach (Lazear 2004)

III. Data

IV. Empirical Results

V. Conclusion

Outline



Output of the employee at firm i: BAy iii )1( λλ −+=

• All skills are of general nature.• These general skills are combined differently in different firms .• A, B = General Skills A and B• λi = Relative Weight of the skills A and B in firm i.• yi = Output at firm i.


• Starting Point: The employee has to decide to what extent he acquiresthe skills A and B in order to maximize his income.


Output of the employee at firm i: BAy iii )1( λλ −+=

• All skills are of general nature.• These general skills are combined differently in different firms .• A, B = General Skills A and B• λi = Relative Weight of the skills A and B in firm i.• yi = Output at firm i.


Output of the employee at firm i

“ outside (expected)

BAy iii )1( λλ −+=(with A>B)

BAy j )1( λλ −+=

λλ −iLevel of specificity


Output of the employee at firm i

“ outside (expected)

BAy iii )1( λλ −+=(with A>B)

BAy j )1( λλ −+=

λλ −iLevel of specificity

Worker Mobility

H1 The more specific the qualification requirements of an occupation(Ausbildungsberuf), the lower is the probability of an occupationalchange after apprenticeship training.

Firm`s Training Investment Share

H2 The more specific the qualification requirements of an occupation, thelarger is the investment of the firm in apprenticeship training.

Hypotheses


BIBB/IAB - Qualification and Career Survey 1979, 1991/92 und 1998/99

Characteristics

• Representative random samples of the German workforce.• Detailed information about skill profiles of the employees.• Information about the vocational carreres of the employees.• Information about training activities of the employees.

Selection

• Persons aged between 16 and 65 years• No civil servants• No self-employed

III. Data


Source: BIBB/IAB-Survey 1979, own calculations.

IV. Explanatory Variable: Specificity Degree

Qualification Portfolio of a Banker

0

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Relative Frequency

Importance of Qualifications

Qualifications



0

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Relative Frequency


Qualifications




German Qualification Portfolio


0

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Relative Frequency


Qualifications

0

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Relative Frequency


Qualifications





0

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Relative Frequency


QualificationsGermany Banker




0

1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Relative Frequency



0

5

10

15

20

25

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Rank

Rank of Qualifications




Determinants of Mobility (during working life)

V. Specificity of Occupation and Mobility

Specificity of Occupation -.0019132 **(.0007577)

male .001205(.0069017)

age .0395193 ***(.0019426)

age2 -.0003772 ***(.0000242)

vocational training .0869616 ***(.0066178)

further controls

N 25.973Wald chi2(14) 1766.39Prob > chi2 0.0000Pseudo R2 0.0508Log pseudolikelihood -16794.486

Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Datasets 1979, 1991/92 und 1998/99. Further controls:groesselehrbetrieb49, groesselehrbetrieb99, groesselehrbetrieb999, groesselehrbetrieb1000, gemeindegroesse100, gemeindegroesse500, gemeindegroesseueb500, welle79, welle91. Level of significance ***=1%, **=5%, *=10%. All estimates of the probit-Regression are estimatedwith robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.


Determinants of Mobility (directly after completion of the apprenticeship training)

Specificity of Occupation -.0007744 ***(.000267)

male.0052922 **(.0024538)

age.0006948 (.0006954)

age2-.0000152 *(8.89e-06)

vocational training.0103447 ***(.0026047)

further controls

N 25.973Wald chi2(14) 681.06Prob > chi2 0.0000Pseudo R2 0.0717Log pseudolikelihood -4470.8469

V. Specificity of Occupation and Mobility

Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Datasets 1979, 1991/92 und 1998/99. Further controls:groesselehrbetrieb49, groesselehrbetrieb99, groesselehrbetrieb999, groesselehrbetrieb1000, gemeindegroesse100, gemeindegroesse500, gemeindegroesseueb500, welle79, welle91. Level of significance ***=1%, **=5%, *=10%. All estimates of the probit-Regression are estimatedwith robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.


Merge of Six Datasets

V. Specificity of Occupation and Firm`s Training Investment


Determinants of Firm`s Net Investment Cost

V. Specificity of Occupation and Firm`s Training Investment

Specificity of Occupation397.2***(119.6)

variance in the skill portfolio 397.3(2592.0)

age at end of apprenticeship training 1967.8***(569.5)

Const. -33 866.3***(12 279.6)

Number of observations 74

F(3, 70) 5.54

Prob > F 0.00

R2 0.19

Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Datasets 1979, 1991/92 und 1998/99 and the BIBB-cost and benefitsstudies of apprenticeship training 1980, 1991and 1998. Level of significance ***=1%, **=5%, *=10%. All estimates of the OLS-regression areestimated with robust standard errors in order to avoide heteroskedasty.


Transfer to Further Vocational Training


Transfer to Further Vocational Training

H1` The more unusual the qualification requirements of an employeein his current firm, the lower is the probability of a job change.

H2` The more unusual the qualification requirements of an employeein his current firm, the larger is the training investment share thefirm has to bear.


V. Specificity and Change of Industry

Specificity of Industry-0.0114***(0.0041)

blue collar0.0166(0.0126)

age0.0516***(0.0075)

age2-0.0005***(0.0001)

exp-0.0032(0.0037)

exp20.0001(0.0001)

male-0.0038(0.0129)

N 12 036

Loglikelihood -7512.2049

Prob> X2 ***

Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Dataset 1998/99. Further controls: level of schooling, level of vocational qualification,income, further characteristics if the employer, further characteristics of the employee. Level of significance ***=1%, **=5%, *=10%. All estimates of theprobit-Regression are estimated with robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.


V. Specificity and Change of Industry (Panel data from the GSOEP)

Specificity of Industry-0.0089***(0.0033)

blue collar0.0134(0.0340)

age0.0010(0.0180)

age2-0.0006**(0.0003)

exp-0.0464***(0.0079)

exp20.0012***0.(0002)

male0.0063(0.0341)

N

Groups48 02712 750

Loglikelihood -10381.053

Prob> X2 ***

Source: Own calculation with STATA 9.2 on the basis of the GSOEP 1996-2003. Further controls: level of schooling, level of vocational qualification,income, further characteristics if the employer, further characteristics of the employee. Level of significance ***=1%, **=5%, *=10%. All estimates of theprobit-Regression are estimated with robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.


V. Individal Specificity and Firm`s Training Investment Share

Indiv. Specificity (Industry) 1.2337***(0.1574)

Indiv. Specificity (Germany) 1.0843***(0.1198)

Distrubution of λ

Lambda 0.0387**(0.0161)

0.0178(0.0127)

Layoff-risk

RiskWiLage 0.0544(0.0426)

0.0567(0.0420)

Market thickness

AMDreg -0.0114***(0.0033)

-0.0115***(0.0033)

AMDbr -0.0045***(0.0009)

-0.0076***(0.0020)

N 13 738 13 738

Loglikelihood

-17881.24

Prob > X2 ***

-17864.13

Prob > X2 ***

Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Dataset 1998/99. Further controls: level of schooling, level of vocational qualification,income, further characteristics if the employer, of the job, of the employee and of the labor market. Level of significance ***=1%, **=5%, *=10%.All estimates of the ZIP-Regression are estimated with robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.


1. The Skill Weights Approach widens the theoretical analysis of importantquestions related to vocational training.

2. The Skill-Weights Approach is very helpful to explain different empiricalfindings consistently:

- The more specific an investment in human capital, the bigger isthe educational investment a firm has to bear.

- The more specific an investment in human capital, the more reduced isan employees mobility after the training.

- These results are robust when estimating on occupational as well as onindustry level.

- The results hold on the level of initial vocational training as well as on thelevel of further vocational training.

VI. Conclusion


3. The apprenticeship training in Germany -formerly seen as general training-is very heterogenous in specificity: Some apprenticeships are more generalwhereas others are highly specific compared to the German labor market.

VI. Conclusion


3. The apprenticeship training in Germany -formerly seen as general training-is very heteogenous in specificity: Some apprenticeships are more generalwhereas others are highly specific compared to the German labor market.

4. Obviously there is a trade-off between firm`s incentive to invest and thespecificity of an apprenticeship training. The higher the specificity,the higher is the firm`s incentive to invest.

VI. Conclusion


3. The apprenticeship training in Germany -formerly seen as general training-is very heteogenous in specificity: Some apprenticeships are more generalwhereas others are highly specific compared to the German labor market.

4. Obviously there is a trade-off between firm`s incentive to invest and thespecificity of an apprenticeship training. The higher the specificity,the higher is the firm`s incentive to invest.

If somebody has the idea to reduce specificity of theapprenticeship training in order to prepare the employeesfor technological changes, mobility might increase. But, at thesame time, the firms would reduce their training investments.

VI. Conclusion


Thank you for your attention!

www.economics-of-education.ch

german apprenticeship training and the skill weights …00000000-5b2d-ab7b-0000... · german...

Documents